Clicky launches dictation that reads your screen before writing

Signals Inbox·July 14, 2026·Voice AI

Clicky has launched dictation that looks at your screen before deciding what to write. Its roughly 450ms speech-to-text is fast, but the bigger move is context: it can see a Claude Code session and form a relevant prompt, or read an email thread and draft a reply in your voice.

Q1What actually launched?

According to Farza’s official announcement, Clicky shipped screen-aware dictation with speech-to-text latency of roughly 450 milliseconds. Instead of only transcribing your words, it looks at the application and content currently onscreen, then uses that context to produce more relevant text.

Q2What does screen-aware dictation actually do?

It reduces how much context you need to say aloud. Inside Claude Code, Clicky can see the code or task onscreen and turn a short spoken request into a fuller prompt. Inside Gmail, it can read the conversation and draft a reply that fits the thread. The screen becomes part of the prompt.

Q3Is Clicky the first to do this?

No. Wispr Flow already uses text near the cursor to improve dictation, while products such as Aqua Voice also advertise screen-aware accuracy. Claude Code now has its own built-in voice mode too. Clicky’s real move is combining fast dictation with the broader screen vision already used by its desktop assistant.

Q4Why does 450ms matter?

Voice input feels broken when users speak, wait, and then fix the result. Around 450ms is short enough for the transcription to feel close to immediate. But speed alone is no longer enough because many products already offer fast speech-to-text. Clicky is betting that understanding the screen will remove more friction than cutting another few milliseconds.

Q5Why is this happening now?

AI users are writing much longer prompts, especially for coding agents. Speaking can be almost three times faster than typing under controlled conditions, and developers are already using tools such as Claude Code and Codex through voice. Wispr Flow reportedly reached a valuation near $700 million, so this is no longer a tiny accessibility niche. Voice is becoming a serious input layer for AI work.

Q6What is the real point of tension?

Clicky is entering a crowded dictation market, so basic transcription will not be enough. The product wins only if screen context saves users from explaining their task and correcting the output. It also creates a privacy trade-off because useful context may include private emails, code, documents, and messages. The next battle is not who hears you best, but who understands what you are looking at without becoming creepy.

Q7So should I care?

Yes, because this is a glimpse of how computer input may change. Today you type instructions and manually paste context. Clicky wants you to speak while the computer gathers that context itself. If it works reliably, dictation stops being a faster keyboard and becomes a simple way to control AI across every application.